Estradiol Metabolism in Cirrhosis

BARNETT ZUMOFF, JACK FISHMAN, T. F. GALLAGHER, and LEON HELLMAN From the Division of Neoplastic Medicine and the Institute for Steroid Research, Montefiore Hospital and Medical Center, New York

A B S l R A C T Abnormal estrogen metabolism has cirrhosis, too, is characterized by the reciprocal eteln found in cirrhosis after administration of relationship between decreased 2-hydroxylation intravenous tracers of estradiol-3H to 6 patients and increased 16a-hydroxylation previously de- and 23 healthy controls. The major abnormalities scribed in hypothyroidism and male breast cancer. observed involved estrogen metabolites other than However, unlike these latter, the increase of 16a- the 3 "classic" ones, i.e., estrone (El), estradiol hydroxy metabolites was less than the decrease (E2), and estriol (E3). Urinary recovery of of 2-hydroxy metabolites. The data indicate clear- radioactivity was regularly elevated in the patients, cut impairment of 2-hydroxylation, suggestive im- to an average of 71 % of the dose compared to pairment of 16a-hydroxylation, and a definite 51% in Tormals. This is considered to reflect the dlelression of the reaction 16a-hydroxyestrone--> component of in trahep-atic cholestasis in cirrhosis. estriol, the latter finding so far unique to cirrhosis. The per cent (lose recovered as urinary glutco- Demonstration of abnormal peripheral metabolism siduronates (42% ) was normal in cirrhotics in of estrogen in cirrhosis provides a new approach contrast to impaired glucuronidation of cortisol to the origin of the hyperestrogenic syndrome in metabolites in this disease. El and E2 were present this disease. in normial anmouints, and E3 was slightly elevated to 21 % of the extract compared to 14%o in con- INTRODUCTION trols. There were strikingly decreased excretion of 2-hydroxyestronie (3% compared with normal The central role of the liver in the transformation 20'j1 ) anid 2-methoxyestrone (2% compared with of steroids and the well-recognized occurrence of 5'/, ) and(l increase(l excretion of 16a-hydroxy- a hyperestrogenic syndrome in liver disease (2) estroiie (12%1o compare(l with normal 6%). Thus hlave stimulated investigation of estrogen metabolism in liver disease. The current status has been A preliminary report of this work has been published summarized in a recent review Schedl in abstract form (1). by (3). Address requests for reprints to Dr. Barnett Zumoff, Virtually all the available data have been obtained Division of Neoplastic Medicine, Montefiore Hospital and either by bioassay or by chemical assay methods Medical Center, New York 10467. limited to the 3 "classical" estrogen metabolites,' Recei-ved for publication 30 June 1967 and in revised El (estrone)l E2 (estradiol), and E3 (estriol). formn 17 August 1967. The following trivial namiies have been used for metabolites: estradiol (F2) :' 3"''°'estratriene-3,17,7-diol estrone (El ) 3-hydroxy-"' '3 (10)estratrien- 1 7-one estriol (E3) A""3 1"'estratriene-3,166a,17,8-triol 16-epiestriol (EF3) At"06'0'estratriene-3,16p,17,8-triol l6a-hydroxyestrone ( 16a-01I- El 3,16a-dillydroxy-A"" 35(0)estratrien-17-one 1 6-ketoestradiol ( 16-0-E-l ) 3,1 7,-dillydroxy-Al"3'510estratrien-16-one 2-hydroxyestronle (2-OH-El) 2,3-dihydroxy--A" 51()estratrien-17-one3. 2-methoxyestrone (2-MeO-E1 2-methoxy-3-hydroxy-A1""'50'0)estratrien-17-onq (20 The Journal of Clinical Investigation Volume 47 1968 The inadequacies of both methods have been well METHODS understood by workers in the field. In addition, Five men, aged 39-50, and one woman, aged 37, were studies utilizing chemical methods have been studied. Each patient had typical clinical and laboratory shown to be subject to two serious limitations. findings of cirrhosis (Table I), as well as a long history together have of alcoholism. At the time of study, each patient was hos- First, the three classical metal)olites had a of the urinary pitalized at the Clinical Research Center and had been shown to constitute about 40%o period of rest, adequate diet, and enforced abstinence from estrogens (4). Other identified metabolites, e.g. alcohol. 2-hydroxyestrone, 2-methoxyestrone, 16-epiestriol, Each subject studied received approximately 1 ,uc of 16a-hydroxyestrone and 16-ketoestradiol, make up estradiol-6,7-'H by intravenous injection of a weighed most of the remainder. Studies that do not include amount of a solution of the steroid in freshly redistilled pyrogen-free propylene glycol. A weighed aliquot of these compounds are incomplete and may, on occa- propylene glycol solution from which the individual sion, be misleading. It was found in the present tracer doses were obtained was used for calculation of the study that an analysis restricted to the three clas- administered dose of radioactivity. Complete urine col- sical metabolites would have yielded a quite er- lections as judged from the consistency of creatinine measurements were obtained for 3 days after each injection. roneous picture of estrogen metabolism in cir- The specimens were refrigerated immediately after void- rhosis. Second, studies of the "urinary production ing and at all times until the collection was complete. rate" of estradiol have shown discrepancies be- If not processed immediately, and this was done in most tween values determined from different metabolites instances, the entire sample was frozen and kept at - 150C of formation until it was possible to separate the steroid metabolites. (5). This indicates that the pathways The radioactive steroids used were homogeneous by iso- of different estrogen metabolites may include vary- topic dilution analyis. The carrier steroids used were re- ing proportions of multiple precursors, some of crystallized to constant melting point before addition which may never have been effective estrogens. to the extract. Purity of carrier steroids was also checked It appears, therefore, that critical evaluation of by thin-layer chromatography and by infrared spec- trometry. All solvents employed were redistilled and of the fate of biologically active estrogen is best ac- high quality. complished by studies of estradiol biotransforma- In each study the 3 day urine collections were combined tion, with radioactive tracers. Very few studies of and divided into 10, 50, and 40% aliquots, of which the this type have been reported. A preliminary com- last was frozen and retained. A known amount (approxi- munication by Crowell, Ereii, and Preedy (6) mately 30 mg) of carrier 2-hydroxyestrone was added to the 10% aliquot; no carrier was added to the 50%o reported' nornmal estradiol l)roduction in cirrhosis, aliquot. Each was incubated with 300 U of 8-glucuroni- but gave few details concerning the pathways of dase2 per ml of urine at pH 5 with acetate buffer at estrogen metabolism. A preliminary communica- 380C for 5 days. The urine at pH 5 was extracted con- tion by Stoa and Thorsen (7) reported little devia- tinuously with ether for 48 hr. The ether extract was metabolism of washed with 9% sodium bicarbonate solution saturated tion from normal in the estradiol with sodium chloride, with saturated sodium chloride solu- cirrhotic patients. In a somewhat more detailed tion, and finally with a very small amount of water, and preliminary communication, Shaver, Roginsky, the solvent was removed. The residue is designated "glu- and Christy (8) reported evidence that suggested cosiduronate extract." It should be noted that sulfate con- an elevated blood estrogen level and delayed jugates are not hydrolyzed by this procedure although 2-deoxy-2-acetamido-D-glucosides (if present) may be; hepatic estrogen inactivation in cirrhosis. double conjugates with glucuronic and sulfuric acid would In the present study the metabolism of tracer not be cleaved to lipid-soluble compounds. These, there- doses of radioactive estradiol was studied in six fore, would be included as "nonglucosiduronate." patients with cirrhosis and was found to be con- The extract-from the 10% aliquot was used only for sistently abnormal, in total excretion, conjugation, measurement of 2-hydroxyestrone and estradiol. A known amount of estradiol-17,8 approximately equal to the and the relative quantities of metabolites. De- amount of 2-hydroxyestrone was added. The mixture was creased efficiency of certain pathways of peripheral acetylated by standing overnight in pyridine and acetic estradiol metabolism was demonstrated, and there anhydride. Excess reagent was removed under a nitrogen formation of a normally minor stream on a water bath, and the acetates were chromato- was increased with 3: 2 metabolite, 16a-hydroxyestrone. The existence of graphed on 8 g of acid-washed alumina. Elution abnormal peripheral estrogen metabolism in cir- 2p.-Glucuronidase, known as Ketodase, was obtained rhosis provides a new approach to the understand- from Warner-Chilcott Laboratories, a division of Warner- ing of the hyperestrogenic syndrome in this disease. Lambert Pharmaceutical Company, New York. Estradiol Metabolism in Cirrhosis 21 TABLE I Clinical Description of Cirrhotic Pattents

Glutamic- Alkaline oxaloacetic plhosphatase Per cent Scrumn transaminase (Bessey- free Dcgree of bilirubin (Karman- Lowry units) cholesterol illness Portacaval (normal Ladue units) (normal Total (normal Subject Age Sex (() 5)* shunt Ascites <0.8) (normal <40) <3.0) cholesterol <34) mg/100 ml mg/100 ml BO 50 M 3 No No 0.3 76 4.2 285 38 NS 42 M 3 No No 1.0 30 4.6 169 39 IN 49 M 1 No No 0.9 36 2.4 235 30 LY 37 F 3 No No 0.3 20 4.2 191 34 HS 39 M 4 Yes No 2.0 59 3.5 230 40 CY 41 M 3 No Yes 2.0 69 3.9 217 40

* 1 = good health; 5 = critically ill.

(v/v) benzene-petroleum ether (boiling range 58'-62'C) Since these two metabolites are not separable by routine gave estradiol diacetate, which was recrystallized to con- chromatographic procedures, the amount of 16a-hydroxy- stant specific activity from petroleum ether-acetone. estrone was determined by reverse isotope dilution. Ali- Elution with 1: 9 (v/v) ethyl ether-benzene. gave 2-hy- quots of the radioactive peak were diluted with unlabeled droxyestrone diacetate, which was recrystallized to con- carrier 16a-hydroxyestrone and then acetylated overnight stant specific activity from petroleum ether-acetone. The with acetic anhydride and pyridine. Purification was ac- total radioactivity of each compound was then calculated complished by preparative thin-layed chromatography on from the specific activity and the weight of carrier. The silica gel, employing successive development in the sys- values are expressed as the free steroid. tems 30% ethyl acetate: cyclohexane and 50% ethyl ace- One-half of the glucosiduronate extract from the 50% tate: cyclohexane. The separated 16a-hydroxyestrone ace- aliquot was submitted to gradient elution partition chro- tate was then crystallized to constant specific activity from matography on acid-washed celite column as described by acetone: petroleum ether. Engel et al. (9). This procedure yields well-defined and In subject CY, carrier 2-hydroxyestrone was added homogenous peaks for 2-methoxyestrone, estrone, 16-epi- to the urine which was processed in the usual manner. estriol, and estriol, and the values for these compounds The "glucosiduronate extract" was separated by counter- were directly determined by combining the pertinent frac- current distribution, as described previously (10). Es- tions and measurement of radioactivity. The peak con- tradiol and 2-hydroxyestrone were separated from the ap- taining 16a-hydroxyestrone also contains 16-ketoestradiol. propriate tubes after carrier addition by acetylation and chromatography, as above. The radioactivity present as estrone, estriol, and 2-methoxyestrone was calculated from TABLE II the contents of the peak tubes. 16a-Hydroxyestrone, 16- Recovery of Radioactivity after i.v. Estradiol-3H ketoestradiol, and 16-epiestriol are recovered together as one area in the countercurrent distribution. The amount % Dose as of 16a-hydroxyestrone present in this study of CY was % Dose nonglucosi- as glucosiduronate determined by reverse isotope dilution, as above. Subject % Dose in urine duronate conjugate Radioactivity was measured in a Packard Liquid Scin- tillation Spectrometer with appropriate corrections for 72 hr st day 2nd + quenching. Specific activities were the mean of at least 3rd day BO 57 48 9 38 19 triplicate measurements and were accurate within the NS 67 54 13 45 22 counting error of ± 5%. IN 66 40 26 37 29 LY 58 44 14 30 28 RESULTS HS 84 70 14 48 36 CY 89 77 12 52 37 Recovery of radioactivity (Table II). The Average of cir- 3 day total urinary excretion of radioactive metab- rhotic patients 71 56 15 42 28 Average of 11 olites was elevated to 71o% of the dose compared normal men 51 28 23 40 11 with 51%o in normal men and 63%o in normal Rasnge (41-77) (17-53) (8-31) (32-51) (1-28) Average of 12 women. This increase was confined to the first normal women 63 39 24 42 21 Range (40-79) (26-59) (14-39) (22-54) (13-31) 24 hr, 56%o of the dose compared with 28%o in normal men and 39%o in normal women. (Women 22 B. Zumof, I. Fishman, T. F. Gallagher, and L. Hellman TABLE II I Estradiol-'H Metabolites in the Glucosiduronate Extract

Per cent of extract 16a-OII-Et 2-oxygenated inetabolites + + Subject ElI E2 E3 2-Meo-El 2-011-El l6a-011-E4,I EE1413 E13 16a-lhydroxy ietabolites

BO 8 5 21 1 1 17 4 38 40 NS 15 6 24 3 4 1(0 5 34 41 IN 15 6 29 1 5 7 5 36 42 LY 5 5 19 1 1 15 2 34 36 HS 26 6 14 4 7 15 4 29 40 CY 23 9 20 1 2 7 * 27 30 Average of cirrhotic 15 6 21 2 3 12 4 33 38 patients Average of normal 20 9 14 5 20 6 5 20 45 patients

* Not determined. normally excrete a larger fraction of the radioac- present would not change the total recovery of this tive estradiol metabolites into urine, both 1st day metabolite to any significant degree. and 72 hr total, then do men ( 11 ).) Excretion on DISCUSSION the 2nd + 3rd days was somewhat less than normal, 15% compared with 24% in normal subjects Estradiol metabolism of the cirrhotic patients was of either sex. The 72 hr recovery of radioactivity abnormal in several respects. An increased fraction in the glucosiduronate extract was essentially of the estradiol metabolites was recovered in urine, identical in the normals and cirrhotic patients. as a result of greater 1st day excretion, with the Thus, by difference, the cirrhotics excreted an in- 2nd + 3rd day significantly below normal. The creased proportion of the dose as nonglucosiduron- normal course of estrogen metabolite (lisposition ate conjugates. These latter were nearly twice as offers a possible exl)lanation for these findings. high in cirrhotics as in the combined normal group, Studies with i.v. estra(liol tracers in patients with and nearly three times as high as in normal men. complete biliary fistulas (13) have shown quanti- Individual metabolites (Table III). The sum tative recovery of radioactivity within 24 hr, about of the seven analyzed metabolites was normal but half in bile and half in urine. This has led to the the pattern was abnormal. 2-Hydroxyestrone and conclusion that in intact subjects the 1st day's 2-methoxyestrone were formed in 14 the normal urinary excretion comprises chiefly metabolites amounts; estriol formation was 50% higher than that have not undergone biliary excretion, whereas normal, and 16a-hydroxyestrone was about twice subsequent collections contain metabolites that normal. The other three metabolites, have undergone prior enterohepatic circulation. It estradiol, would be anticipated that a decrease in hepato- estrone, and were present in normal 16-epiestriol, biliary excretory capacity, a regular pathophysio- amounts. logical feature of cirrhosis corresponding to the In subjects IN and NS the "nonglucosiduron- anatomically demonstrable intrahepatic cholestasis ate" conjugates after ketodase treatment were (14), would cause increased 1st day urinary elim- solvolyzed (12), and the metabolites were sep- ination of estrogen metabolites. This is analogous arated by the described procedures. It was found to the increase of urinary bilirubin when hepato- that 2-methoxyestrone and 16a-hydroxyestrone biliary excretory capacity is reduced in disease. In represented almost exactly the same per cent of this agreement with this explanation, the highest 1st extract as they did of the "glucosiduronate con- day recovery of estrogen metabolites occurred in jugates." Estriol represented a somewhat smaller the two patients with the greatest elevation of se- per cent of the solvolyzed extract but the amount rum bilirubin concentration. The decreased 2nd + Estradiol Metabolism in Cirrhosis 23 OH o major pathways varies in disease but the sum of the two appears to be constant in all diseases HO )@ HO v HO studied so far. D)epression of one leadls to eqniv- ESTRADIOL ESTRONE 16- EPIESTRIOL aleiit accetlwation of the otlher.

0 0 In cirrhosis, the decrease of 2-oxygenated metabolite formation was profound but there was only HO) HO N a small increase in estriol formation. In this sense 2-HYDROXYESTRONE 160C-HYDROXYESTRONE the formation of estriol was relatively depressed, I ¶0 OH since the increase was not consistent with the decrease in 2-oxygenated metabolites. Even when HO HO 2- METHOXYESTRONE ESTRIOL the 16a-hydroxyestrone excretion was considered, FIGURE 1 Normal pathways of estradiol metabolism in the increase in total 16a-hydroxy metabolite forma- man. tion was only about % as great as the decrease in 2-oxygenated metabolite formation. This con- 3rd day recovery in the cirrhotic patients is a trasts with hypothyroidism (17) and male breast result of the large urinary loss on the 1st day, cancer (10) where the increase in 16a-hydroxy since less of the estrogen metabolites reach the in- metabolites is equal to or slightly greater than the testine to be available for reabsorption. decrease in 2-oxygenated metabolites. These data In contrast to the elevated total urinary recovery suggest some impairment of the 16a-hydroxylation The ratio of 16a-hydroxy- of estrogen metabolites, the proportion excreted as reaction in cirrhosis. conjugates was normal. This dif- estrone to estriol in the cirrhotic patients was ap- glucosiduronate twice normal and approximately five from metabolism in cirrhosis, where proximately fers cortisol times as high as in male breast cancer. These decreased glucosiduronate conjugation of metab- the The results indicate inhibition or partial loss of olites was observed (15). glucosiduronate for reduction of fraction showed a normal total amount of the dehydrogenase responsible the 16a-hydroxyestrone to estriol. seven analyzed metabolites, with an abnormal pattern. There was a decrease of the 2-oxygenated It is tempting to relate the observed abnormal- metabolites, 2-hydroxyestrone and 2-methoxy- ities of peripheral estrogen metabolism to the estrone, to % of normal, with an increase of the genesis of the hyperestrogenic syndrome in cir- 16a-hydroxy metabolites. Estriol was 50% higher rhosis. An increased hormonal effect is presumably than normal, and 16a-hydroxyestrone was present due to elevated blood (and/or tissue) concentra- in amounts more than twice normal. These changes tions of active hormone. This in turn could be due in estrogen metabolites do not appear to be a con- to increased hormone production, decreased inac- sequence of the altered conjugation mechanisms in tivation, or both. Studies of radioactive estradiol cirrhosis since analysis of the "sulfate" fraction in metabolism by Shaver. Roginskv. and Christy (8) two patients disclosed quantities of 2-methoxy- led these authors to the conclusion that blood estrone and 16a-hydroxyestrone relatively similar estrogen levels were indeed elevated in cirrhosis. to that present in the "glucosiduronate" extract. They concluded also that the elevated blood levels In order to interpret these changes, it is helpful occurred without elevated estrogen production, cit- to consider the over-all pattern of estradiol metab- ing unpublished data of their own and a report by olism (Fig. 1) (16). Estradiol forms a reversible Crowell, Eren, and Preedy (6) which indicated redox system with estrone, and all the other prod- normal estradiol production in cirrhosis, and they ucts are derived from estrone. The latter is metab- speculated that the initiating cause of the hyper- olized along two major pathways, the 2-oxygenation estrogenic syndrome in cirrhosis is impaired he- pathway (estrone -* 2-hydroxyestrone -> 2-meth- patic removal of estrogen. oxyestrone) and the 16a-hydroxylation pathway As Shaver et al. pointed out, their studies pro- (estrone -* 16a-hydroxyestrone -> estriol). For- vided no details concerning the possible step(s) mation of 16-epiestriol represents a separate minor of hepatic estrogen metabolism which might be pathway. The relative prominence of the two impaired. The present study elucidates but does 24 B. Zumoff, J. Fishman, T. F. Gallagher, and L. Hellman not solve this problem. One of the two major 6. Crowell, G. C., S. Eren, and J. R. K. Preedy. 1963. metabolic pathways that inactivate estrone, i.e. 2- Further studies of estrogen secretion rates in normal adults and males with chronic liver disease. In Pro- oxygenation, is clearly impaired. The status of the ceedings of the 45th Meeting of the Endocrine So- other pathway, 16a-hydroxylation, is uncertain. ciety, Atlantic City, N. J. J. B. Lippincott Co., Phila- Of its two sequential steps, the second is definitely delphia. 44. impaired. The evidence concerning inhibition of the 7. St6a, K. F., and T. Thorsen. 1966. Excretion of free first step is suggestive but not conclusive. and conjugated metabolites in the urine of normal in- dividuals and patients with liver cirrhosis following The role of the increased amounts of 16a-hy- intravenous administration of oestradiol-17p8-"C. Ex- droxyestrone requires evaluation. If this metabolite cerpta Medica International Congress Series. Second were estrogenic, an explanation for the hyperestro- International Congress on Hormonal Steroids. 111: 311. (Abstr.) genic syndrome would be at hand. From animal 8. Shaver, J. C., M. S. Roginsky, and N. P. Christy. assays (18) the estrogenicity of 16a-hydroxy- 1963. Clearance from plasma of isotopically labelled estrone, like that of estriol, is weak. This conclu- oestradiol in patients with hepatic cirrhosis. Lancet. sion is not necessarily transferrable to man. Direct 2: 335. study of the human biological effects of 16a- 9. Engel, L. L., C. B. Cameron, A. Stofyn, J. A. Alex- ander, 0. Klein, and N. D. Trofimow. 1961. The hydroxyestrone, its estrogenicity as well as the estimation of urinary metabolites of administered possibility that it may inhibit 2-hydroxylation and/ estradiol-47#-16-C". Anal. Biochem. 2: 114. or 16c-hydroxylation of estrone, seems indicated. 10. Zumoff, B., J. Fishman, J. Cassouto, L. Hellman, and T. F. Gallagher. 1966. Estradiol transformation in ACKNOWLEDGMENTS men with breast cancer. J. Clin. Endocrinol. Metab. 26: 960. The authors gratefully acknowledge the invaluable tech- 11. Zumoff, B., J. Fishman, J. Cassouto, T. F. Gallagher, nical assistance of Mr. Morris Kirschenbaum and Misses and L. Hellman. 1968. The influence of age and sex Ruth Jandorek, Gertrude Gilman, and Rosemarie Lehman. on normal estradiol metabolism. J. Clin. Endocrinol. This work was supported in part by a grant from the Metab. In press. American Cancer Society and by Grants CA-07304 from 12. Fukushima, D. K., T. F. Gallagher, W. Greenberg, the National Cancer Institute and FR-53 from the Gen- and 0. H. Pearson. 1960. Studies with an adrenal eral Clinical Research Centers Branch, National In- inhibitor in adrenal carcinoma. J. Clin. Endocrinol. stitutes of Health, Bethesda, Md. and Metab. 20: 1234. 13. Sandberg, A. A., and W. R. Slaunwhite, Jr. 1957. REFERENCES Studies on phenolic steroids in human subjects. II. The metabolic fate and hepato-biliary-enteric circula- 1. Zumoff, B., J. Fishman, T. F. Gallagher, and L. tion of C14-estrone and C"-estradiol in women. J. Clin. Hellman. 1967. Abnormal estrogen metabolism in Invest. 36: 1266. cirrhosis. J. Clin. Invest 46: 1136. (Abstr.) 14. Popper, H., F. Schaffner, E. Rubin, T. Barka, and F. 2. Barr, R. W., and S. C. Sommers. 1957. Endocrine Paronetto. 1963. Mechanisms of intrahepatic cholesta- abnormalities accompanying hepatic cirrhosis and sis in drug-induced hepatic injury. Ann. N. Y. Acad. hepatoma. J. Clin. Endocrinol. Metab. 17: 1017. Sci. 104: 988. 3. Schedl, H. P. 1965. Steroid hormone metabolism in 15. Zumoff, B., H. L. Bradlow, T. F. Gallagher, and L. liver disease. In Progress in Liver Disease. H. Popper Hellman. 1967. Cortisol metabolism in cirrhosis. J. and F. Schaffner, editors. Grune and Stratton, New Clin. Invest. 46: 1735. York. 2nd edition. 104. 16. Fishman, J., L. Hellman, B. Zumoff, and J. Cassouto. 4. Fishman, J., J. B. Brown, L. Hellman, B. Zumoff, 1966. Pathway and stereochemistry of the formation and T. F. Gallagher. 1962. Estrogen metabolism in of estriols in man. Biochemistry. 5: 1789. normal and pregnant women. J. Biol. Chem. 237: 1489. 17. Fishman, J., L. Hellman, B. Zumoff, and T. F. Gal- 5. Gallagher, T. F., J. Fishman, B. Zumoff, J. Cassouto, lagher. 1965. Effect of thyroid on hydroxylation of and L. Hellman. 1967. Studies of production and estrogen in man. J. Clin. Endocrinol. Metab. 25: 365. metabolism of estrogens in human breast cancer. In 18. Loraine, J. A. 1958. Oestrogens. In The Clinical Ap- Endogenous Factors Influencing Host-Tumor Balance. plication of Hormone Assay. E. and S. Livingstone, University of Chicago Press, Chicago. 127. Ltd., Edinburgh. 154.

Estradiol Metabolism in Cirrhosis 25